Donald Trump has said that he will not run as a third-party candidate, but Republican leaders are no doubt still considering the possibility. They’ve also met to discuss the possibility of Trump-induced chaos at the Republican National Convention, and have considered how to prevent Trump from winning the nomination. But there’s a bigger question lurking here: Which would be worse for the party, a Trump nomination or a third-party candidacy?
For Republicans, both scenarios seem full of potential disaster. Third-party runs conjure images of divided voting blocs and general-election losses — think Theodore Roosevelt in 1912 (whose third-party candidacy helped cost Republicans the White House), Ross Perot in 1992, or even Ralph Nader in 2000. But nominating Trump hardly seems like a good alternative for the party. High-level party leaders, including House Speaker Paul Ryan, have already denounced Trump’s brand of xenophobia, and most political observers think Trump would be a long shot in the general election and could hurt the chances of Republicans in congressional and state races.
Both outcomes would probably be bad for the GOP, but a Trump third-party run would be worse.
It’s not hard to see why party leaders would be concerned with the possibility of nominating Trump in the short term. Conventional wisdom holds that ideologically extreme candidates suffer an “electoral penalty.” This phenomenon is supported by a lot of evidence, but it’s not entirely clear how it applies to Trump. His positions on immigration are extreme. But as Josh Barro at The New York Times pointed out, Trump’s stances on other economic and social issues hardly fit the profile of an orthodox ultraconservative. In other words, it’s difficult to predict the impact of an unconventional candidate like Trump on the national vote.
And maybe Trump wouldn’t have much effect at all: Political scientists have been saying for some time that candidate gaffes and other day-to-day campaign developments are often overblown, with much of the national two-party vote determined by broader factors like the state of the economy and the duration that the incumbent party has held the White House. And remember, we’re talking about a world in which Trump has won the Republican nomination — he would already have won over a good portion of the GOP.
So it’s not at all clear a Trump nomination would affect how people vote, but what about who will vote? There are lots of theories and studies about why turnout varies by age, ethnicity and other characteristics, or how various campaign techniques can encourage people to vote. But the study of aggregate voter turnout tends to focus on cross-country comparisons; comparing across U.S. elections over time is a bit trickier — few observations, many complex and interrelated variables. Nonetheless, drawing on what we know about Trump supporters and about voting behavior, we can still surmise that a Trump nomination would do less damage than a third-party run.
Probably the most important thing here is to emphasize that, despite conventional wisdom, candidates probably have marginal effects on aggregate turnout. Scholars have attributed the decline in voter turnout after the 1960s to changing social structures, voter apathy and disengagement, changing campaign tactics and a decline in the percentage of eligible voters among the over-18 population. Nevertheless, there’s substantial variation from year to year. Turnout has been high in the past few presidential-election years, when the races have been close and the two parties’ policy programs have been distinct, while turnout hit a low point in 1996, when few people considered the race competitive.
So what can we predict about voter turnout in the event of a Trump nomination, or a three-way race? Three-way contests haven’t led to higher turnout on a consistent basis.1 But will Trump voters stay home if Bush, Cruz, Rubio, etc. is the nominee? Or will supporters of those candidates sit out a Trump candidacy?
Based on what we know about voter turnout, these two scenarios are not the same thing. Supporters of Bush, Cruz, Rubio or other mainstream candidates are much more likely to be reliable voters. Iowa State political scientists David Andersen and David Peterson found, for example, that Trump supporters in Iowa are more likely to have spotty records of participating in primaries. It’s possible that these findings don’t directly apply to general election participation in the rest of the country, but the evidence doesn’t end there. Two researchers, Doug Ahler and David Broockman, also found that Trump supporters tend to fit the profile of a “moderate” voter — which is to say, inconsistent. In this Monkey Cage piece, they presented evidence that while conservatism on immigration is, predictably, correlated with Trump support, so is liberalism on taxes. Ideological consistency is strongly associated with turning out to vote (see Alan Abramowitz’s “The Disappearing Center”), as is past voting.
All this is to say that Trump voters are more likely to defect from the Republican Party if one of the other candidates gets the party’s nod — either to stay home or to support the third-party ticket. Reliable Republican voters, on the other hand, are a better bet to show up and support the party ticket, even if they have to hold their noses to do so. The data we have about how turnout works suggests that the GOP will lose fewer voters to a Trump nomination than to a third-party candidacy.
What about the lasting impact of a Trump nomination on the Republican Party brand? Commentators — with either concern or glee — have noted that the Trump candidacy could spell doom for the Republican Party in the long haul. But does evidence support this?
Not really. Failed presidential candidates don’t drag their parties down as much as we might guess. Turning to recent history, the evidence of lasting impact is pretty scant. Barry Goldwater lost big in 1964, but the Republican Party rebounded and won the White House four years later. Similarly, George McGovern did much worse than the economy and other factors would have predicted in 1972, but the Democrats won the next presidential election (admittedly, with the help of Watergate). These losses also had opposite impacts on their respective parties’ ideologies. The Goldwater candidacy is credited with injecting movement conservatism into presidential politics, even though it was unsuccessful in the short term. McGovern’s candidacy, on the other hand, precipitated a move to the center, resulting in Carter’s 1976 candidacy and a general shift away from mid-20th-century liberalism.
The impact of an ideologically charged, unsuccessful presidential candidacy seems to depend on larger, long-term factors. In the 1960s and 1970s, the New Deal coalition was splintering, while the conservative movement was ascendant. This shaped how candidates like Goldwater and McGovern, who drew similar numbers in general elections, affected their parties.
The lessons of history suggest, instead, that significant damage to party reputations is done by unsuccessful presidencies, not unsuccessful presidential candidates. Unsuccessful presidents like Herbert Hoover and Carter shaped their parties’ reputations for decades after (see, for example, attempts to compare Obama to Carter). But Trump’s approach and lack of real party roots probably make him more like an even worse president, Andrew Johnson, whose myopia and racism brought down more than just his party. Republicans stand a smaller chance of electoral loss if they nominate Trump than if he launches a third-party bid. But nominating Trump might be the outcome that should worry party leaders the most. Trump winning the nomination, and then winning the presidency — as unlikely as that may be — probably represents the greatest long-term risk to the Republican Party.